Issues with open many browsers with uc_gui_handle_cf and ThreadPoolExecutor

[SenseiSol]

Here's the code:
def parse_top_traders_in_threads(top_tokens):
driver = Driver(undetected=True)
parser = GmgnParser(driver)
try:
for top_token in top_tokens:
parser.parse_top_traders(top_token)

except Exception as ex:
logger.error(f"Ошибка при обработке токена {top_token}: {ex}")
finally:
driver.quit()
def main(threads_qty):
top_tokens = parting(parse_tokens_from_excel(), threads_qty)
with ThreadPoolExecutor(max_workers=threads_qty) as executor:
for _ in range(threads_qty):
executor.map(parse_top_traders_in_threads, top_tokens)

if name == 'main':
main(10)

using for json parse(bypass cf challenge)

def init(self, driver):
self.driver = driver
self.driver.maximize_window()

def get_json(self, url):
self.driver.uc_open_with_reconnect(url)
self.driver.uc_gui_click_cf()

# Получаем HTML-код страницы
html_content = self.driver.get_page_source()
soup = BeautifulSoup(html_content, 'html.parser')

# Ищем элемент <pre> для получения JSON
pre_element = soup.find('pre')
if pre_element:
    try:
        return json.loads(pre_element.text)
    except json.JSONDecodeError as e:
        logger.error(f"Ошибка декодирования JSON: {e}. Ответ: {pre_element.text}")
        return None
else:
    logger.error("Не удалось найти элемент <pre> на странице.")
    return None

after 1-2 hours working I always had errors like:
Target window is already closed
Can't open DEV Tools and more

I think it is because of cf bypass, because it is using GUI for
what is the best way to run the program?
I need to get data from about 30k pages

Can I use few Process instead of single?

Ran my program at windows server with 42 threads support

[mdmintz]

For multithreading, you have two options:

1. pytest via pytest-xdist. (Eg. pytest -n3, as seen in SeleniumBase/examples/presenter/multi_uc.py)
2. ThreadPoolExecutor as shown in SeleniumBase/help_docs/uc_mode.md.

Note that now there's CDP Mode, which is an upgrade over regular UC Mode. Add sb.uc_gui_click_captcha() as needed.

The number of threads should not exceed the number of CPUs on a system.

Also note that the CAPTCHA-clicking methods use PyAutoGUI: Only one thread is allowed to move the mouse at one time.

[SenseiSol]

Is there other way to bypass cf on seleniumbase?
need to parse like 30k pages